Digital and technology policy and national sovereignty

Approximate read time: 48 minutes

1. Key points

• Digital sovereignty concerns the extent to which nation states can retain control over digital infrastructure, technologies and data in an interconnected world.
• Rapidly developing technologies and the concentration of key digital services in nations such as the US and China have called into question how other nations such as the UK can and should adapt to these cross-border technologies, and how they can ensure their country’s industries and interests are not marginalised in an evolving digital world.
• Concerns have been prompted by several issues, including recent incidents where geopolitical tensions have emerged over how digital services should be taxed, for example, and online protections.
• Advocates of digital sovereignty point to the strategic risks posed by dependence on overseas providers. In contrast, others believe complete digital sovereignty is unrealistic and call for international collaborations and greater use of open technologies.
• The UK has not pursued an explicit strategy of digital and technological sovereignty but rather has sought to develop sovereign capabilities in areas such as AI whilst leveraging international collaboration.
• However, some stakeholders have called upon the UK government to go further in developing a more structured cross-government approach to digital sovereignty.

2. National sovereignty in the digital age

The House of Lords is scheduled to debate the following motion on 23 July 2026:

Baroness Kidron (Crossbench) to move that this House takes note of the impact of digital and technology policy on national sovereignty.

Baroness Kidron is the honorary president, founder and former chair of the 5Rights Foundation, an NGO which campaigns for children’s rights in an online world, a visiting professor in practice at the London School of Economics, where she chairs the Digital Futures for Children research centre, and a member of the Lancet Psychiatry’s Commission on Global Action on Problematic Usage of the Internet.

In the UK and Europe, concerns about the impact of technology and digital policy and their impact on national sovereignty have arisen for several reasons. They include concerns, for example, around the potential security risks posed by Chinese technologies, the market dominance of small numbers of very large and influential companies—notably the so-called ‘big five’ of Microsoft, Apple, Alphabet (Google), Amazon and Meta, all of which are based in the United States—and the rapid evolution of potentially transformative technologies such as artificial intelligence, robotics and quantum computing.

These factors have called into question how nation states such as the UK can and should adapt to these cross-border technologies, and how it can ensure the country’s industries and interests are not marginalised in an evolving digital world.

This briefing examines some of the key thematic areas which have created challenges for national sovereignty, and how the UK government and European nations and institutions have responded to date. First, it examines how digital sovereignty could be defined from infrastructure to information.

3. What is digital sovereignty and what issues is it seeking to address?

3.1 Digital sovereignty in a time of rapid technological evolution

Defining digital sovereignty

Views about how to achieve digital sovereignty differ because there is little agreement on precisely what it looks like or what its key outcomes should be.

A consistent theme, however, is arguably that at a nation-state level digital sovereignty means the capacity of that state to govern the rules, standards, infrastructures and data flows that underpin the digital economy within its territory on its own terms.^[1]

The concept of agency is also linked to this definition. Consultancy firm Public Digital defines agency as the capacity to make a “deliberate decision” rather than being forced down a path by circumstance.^[2] Similarly, it suggests capacity itself is the “institutional competence to choose well”—in other words having the skills, processes, and knowledge, including at leadership level, to act in the digital sphere.

Digital sovereignty is often framed around three layers:^[3]

• Infrastructure/physical layer: This includes the physical infrastructure required for digital services including servers, data centres, networks, cloud and edge infrastructure.
• Code/standards layer: This includes software, algorithms, protocols, design frameworks and technical standards.
• Data/content layer: This includes the data and content used in the digital sphere, and how it is stored, processed, moved, and owned.

Market dominance and geopolitical concerns

Countries can take a different approach to each of these layers, and concerns around digital sovereignty are often complicated by an absence of national providers. For example, in cloud computing, the top three US firms (Amazon Web Services (AWS), Microsoft Azure, and Google Cloud) reportedly accounted for almost two thirds of the global market in 2025.^[4] The only non-US companies in the top eight of global providers (which together accounted for 75% of global services in 2025) are Chinese, as shown in figure 1.

Figure 1. Global cloud computing market share (Q3 2025)

Concern around the market dominance of US cloud providers is driven in part by the powers provided by the Clarifying Lawful Overseas Use of Data (CLOUD) Act 2018. This legislation grants the US government powers to require US-based companies to hand over data they host, even if the data is held outside the US. To counter these fears, several providers such as Amazon, Microsoft and Google claim to offer ‘sovereign’ cloud services in Europe, where they guarantee to customers that their data will be held in European data centres.^[5] However, critics such as Stefan Keller, chief product officer at company Open Systems, contend that such contractual safeguards cannot insulate European data from the reach of extraterritorial US laws. He points to testimony given by Microsoft France’s legal director, Anton Carniaux, to the French Senate:

In June 2025, Microsoft France’s legal director was asked under oath before the French Senate whether he could guarantee that French citizens’ data would never be transferred to US authorities without French consent. His answer was simple: No. That single word confirmed what many European security leaders have been debating for years. Data sovereignty is not a setting you enable. It is not a checkbox in a vendor contract. And it is no longer primarily a legal question. It is an architecture decision, and the window to get it right is closing.^[6]

Civo, a UK-based cloud computing provider, has written about the potential implications of this for the UK in its 2026 digital sovereignty report:

The US CLOUD Act further grants US authorities the power to request access to data held anywhere in the world from providers subject to US jurisdiction, with no legal requirement for providers to inform customers. This is a structural condition of the infrastructure, not a theoretical risk, and it is difficult to assess: only 35% of UK organisations have full clarity on where their data is located and governed, meaning the majority cannot accurately quantify their exposure.^[7]

UK parliamentarians have raised the issue of the US CLOUD Act for government and public sector data held on infrastructure provided by companies such as Microsoft, Google Cloud and AWS. In response to a written parliamentary question in March 2026, Department for Science, Innovation and Technology (DSIT) Minister Ian Murray said:

DSIT has not made any central assessments of the US CLOUD Act and the implications for UK government data.

Under UK data protection laws, UK organisations, including UK government, must ensure personal data is appropriately protected when transferred internationally.

The UK has an adequacy decision for certain transfers to the US under the UK Extension to the EU-US Data Privacy Framework, which included an assessment of US government access laws and practices, such as the US CLOUD Act. Where adequacy is not relied upon, organisations must use alternative safeguards, such as standard contractual clauses, in accordance with Article 46 of the UK GDPR.^[8]

Access fears

Fears have also been expressed that the dominance of US technology providers could theoretically be used to shut off access to key services in other jurisdictions in response to political tensions or concerns. The Open Rights Group, which campaigns to protect individuals’ rights to privacy and free speech online, has called for a UK digital sovereignty strategy. It argues:

Even systems that are technically secure are strategically fragile if a foreign company or power can pull the plug. If the UK got into an argument today with [President] Trump—over Greenland, Israel, or trade—then Trump could swiftly close down the UK’s government, by closing down US owned IT and cloud systems.^[9]

Such concerns were heightened in 2025 when Karim Khan, chief prosecutor at the International Criminal Court (ICC), lost access to his Microsoft Outlook email account after being sanctioned by the US.^[10] In a statement announcing the sanctions, the White House said the ICC had “engaged in illegitimate and baseless actions targeting America and our close ally Israel”.^[11] A Microsoft spokesperson said that the company was in contact with the court “throughout the process that resulted in the disconnection of its sanctioned official from Microsoft services” and that “at no point did Microsoft cease or suspend its services to the ICC”.^[12] Since that time, 11 of the ICC’s officials—including the chief prosecutor and eight judges—have been placed under US sanctions, subjecting them to measures that include bans on travel to the US and fines and prison sentences for American companies that provide them services.^[13] As a consequence, Canadian ICC judge Kimberley Prost, for example, has reportedly had her Google and Amazon accounts cancelled. The Open Rights Group has described such moves as proof that the risks it describes are “not theoretical”.^[14] Others such as Matthias Bauer, director at the European Centre for International Political Economy, have suggested they were “highly sceptical” that the Trump administration would choose to use a digital ‘kill-switch’ to suspend digital services in Europe or the UK.^[15]

Disagreements over digital taxation and online protections

Geopolitical tensions have also followed the imposition of regulation designed to make online platforms more accountable for illegal content, disinformation and child protection. In December 2025, partly in reported response to the EU’s Digital Services Act (DSA), the US State Department imposed a visa ban on five Europeans it claimed had “led organised efforts to coerce American platforms to censor, demonetize, and suppress American viewpoints they oppose”.^[16] They included ex-European Commissioner Thierry Breton, who reportedly played a key role developing the DSA, though Mr Breton himself has also come in for criticism for politicising the DSA.^[17]

The UK’s Online Safety Act (OSA) has provided a similar point of tension. Multiple sources have reported that the US government has been seeking changes to the OSA, and other areas of technology policy such as the UK’s digital services tax, as part of trade negotiations between the US and UK.^[18] Reportedly, such tensions are a factor in the lack of progress in talks on the Technology Prosperity Deal—a multibillion dollar agreement for some of the largest US technology companies to invest in the UK, announced in a memorandum of understanding in September 2025.^[19] A range of investments from US technology companies were announced as part of the agreement, including a $30bn (£22bn) commitment from Microsoft into AI infrastructure and ongoing operations and a £5bn commitment from Google parent company Alphabet.^[20] In December 2025, the US suspended implementation of the Technology Prosperity Deal.^[21] In response to recent parliamentary questions, the UK government has said that ministerial level discussions between the two nations were ongoing:

The United States is our close ally and tech partner, and we are committed to ensuring that bond delivers real benefits for hardworking people on both sides of the Atlantic.

We continue to have ministerial-level discussions with the US administration on science and technology, including on joint initiatives in the Tech Prosperity Deal.^[22]

Similarly, recent debate has focused on opposition from the US government to the European Union imposing a digital services tax on some of America’s biggest technology companies. The prospect of the EU introducing such a tax—which is already in place in the UK, as discussed in section 4.1 below—has led to US President Donald Trump threatening a 100% import tariff on any nation which imposes such a levy.^[23] It is unclear how the UK might be affected by such moves. However, in April 2026, President Trump reportedly threatened to impose a “big tariff” on the UK if it did not drop its digital services tax.^[24]

Location of emerging technologies and capabilities

As emerging technologies such as generative artificial intelligence, robotics and quantum computing continue to evolve at a rapid pace, there is also concern at so many of the leading companies in these areas being predominately based in either the US or China. According to the most recent edition of Forbes magazine’s AI 50 list published in April 2026, which profiles the 50 “most promising” privately held AI companies around the world (which can exclude Chinese companies as a result of their ownership models), 44 of those companies were headquartered in the United States.^[25] This included OpenAI and Anthropic, two “AI giants” which Forbes said had accumulated a combined $242.6bn in venture funding, “about 80 percent of the total $305.6bn that the companies on this year’s AI 50 list have raised”. Sweden had two companies in top 50, whilst Germany, France and Canada all had one. The UK also had one company in the top 50—Synthesia, the AI avatar and video game maker.

Concerns over the ownership of AI models have grown as these models have become more powerful. The United States government temporarily placed export controls on Anthropic’s Fable 5 and Mythos 5 models over apparent fears of their capabilities and citing national security concerns.^[26] OpenAI has faced similar restrictions over its GPT-5.6 model. All the models have now been made available, reportedly subject to restrictions which have been placed on the tools to address security concerns. Fears over the capability of these models, and the ability of the US to decide who has access to them, have led to further questioning of the UK’s reliance on overseas AI providers.^[27]

Public sector providers and ‘vendor lock-in’

The House of Commons Science, Innovation and Technology Committee has also raised the issue that many UK public systems and services are dependent on a small number of large digital and technology providers, such as Palantir, Microsoft and AWS.^[28] The committee argued this could lead to ‘vendor lock-in’, where an over-reliance on external suppliers can create significant obstacles to switching between suppliers, such as the high costs of exiting long-term contracts, technical complexities, reliance on proprietary technologies and embedded idiosyncratic data formats. The committee said the government appeared “worryingly comfortable” with this arrangement, and singled out the US company Palantir in particular as a source of concern:

Of the small number of technology providers that the UK public sector relies upon, Palantir concerns us most. In the United States it has supplied software for that country’s military and immigration services, supporting highly controversial policies and activities. Its co-founder has criticised the concept of a national health service and the company has issued a manifesto that makes explicitly political arguments, undermining what the head of their UK and European business told us. In the UK Palantir has increased its presence across the public sector despite this clear mismatch with UK values. We, however, are of the view that Palantir should not have such a significant role in the UK public sector, and that it is far from the only company capable of providing the data analysis ‘middleware’ required by public bodies.^[29]

The committee also called for the UK to take greater steps towards digital sovereignty which are explored in section 4 below.

How much should foreign ownership be a concern?

Some commentators have suggested that ownership of companies or services is not necessarily a problem but rather what is crucial is making sure that digital services and technologies are effectively policed and regulated regardless of where they originate. Amanda Brock, CEO of Open UK, open tech’s industry body which advocates for open technology, contends that anti-US and anti-tech sentiment can become intertwined, and it is not possible to build a technology stack—that is the range of technologies needed to build or deliver a service—without internationalism:

If we’re honest, much of the aggression directed at the US in this context is thanks to the success of its tech sector. Criticisms are often intertwined with anti-big-tech sentiment. Unsurprisingly, the US administration has come out fighting. These sovereignty-based attacks have directly contributed to the Trump administration’s tariffs.

Every country should seek the best innovation for its citizens. And that is accessed through global collaboration. Tech stacks must be open and visible to diminish dependencies on particular suppliers. Almost all (96%) commercial software is somewhat dependent on open-source technology, which is based on international contributions that have long tails of dependencies on other code. It is impossible to build a valuable stack today without internationalism.^[30]

She argues, as is further explored in section 4.4 below, for more use of open-source technology—namely the practice of publishing digital resources publicly alongside their source code or source files, enabling use, study, modification, and redistribution, rather than relying on proprietary technologies.^[31] Amanda Brock contends:

Governments should focus on opening technology up instead of policing where it’s from. Building a software stack from scratch requires open-source technology, which is built by a collaborative, global community. If technology is transparent, we can iterate and innovate on top of it.^[32]

In contrast, Kevin Dallas, CEO of US open-source software company Enterprise DB, argues that sovereignty around data and AI is “existential”.^[33] He contends:

Applications are rapidly integrating all four types of artificial intelligence: predictive, generative, agentic and, soon, physical. But there’s no AI without data. If you don’t provide the right data, algorithms can hallucinate and deliver false insights. Control over AI systems, as well as the data those systems use, is therefore an existential concern.

What does this mean? If companies are to benefit from AI tools, they must control the data that feeds those tools and be the ones to decide how and where AI is deployed. And the government in their jurisdiction must control the security, compliance and regulatory requirements of the AI used within their boundaries.^[34]

He adds that sovereign AI capability is high on the agenda for many nations as a result, including the UK:

Sovereign AI is a national agenda item for China, the UK, Germany, Saudi Arabia, the UAE, India and so on. These countries realise that they need AI sovereignty to be globally competitive. There’s also a huge appetite for investors and enterprise CEOs to embrace sovereignty and control their future with a sovereign-data-layer platform.^[35]

The UK’s approach to sovereign AI capabilities is examined in more detail in section 4.3 below.

3.2 Different models of digital sovereignty

In response to these factors and constraints, there are several different models of digital sovereignty being pursued by countries around the world. In a 2024 article, Samuele Fratini et al identify the four models which are briefly summarised below, and then in the subsequent figure 2 below Fratini et al categorise certain nations’ approach according to these models. The models are as follows:^[36]

• Rights-based model

The rights-based model is characterised by the framing of digital regulation as a means of strengthening fundamental and democratic rights, the attempt to find a balance between rights and market benefits, and a resolute regulatory approach to private companies. The ultimate objective of this model is the reinforcement and introduction of rights for citizens in the digital space. Asserting user control over data is a key component of this model (eg data protection regulations) and the need to secure fundamental rights can also inform efforts to constrain the power of, and gain independence from, large technology companies.

• Market-orientated model

The market-oriented model values economic freedom, market benefits, and competition-based innovation at the expense of state regulation. Self-regulation through codes of practice rather than legislative or statutory frameworks may be preferred. In practice this model can involve regulatory or fiscal support for technology companies rather than a pure economic laissez-faire approach. Where digital regulation is introduced, it is predominately for reasons such as national security (for example, the decision taken by many Western countries to ban Chinese technology company Huawei from 5G networks).

• Centralisation model

The centralisation model is characterised by the reshaping of digital governance based on centralised regulation—in other words, the exclusion of non-state actors from the process of decision-making. The state is seen as sovereign in the digital world and governments restrict or ban foreign digital companies and favour home-grown (sometimes state-owned) alternatives, which are easier to control by government agencies. States aim for a degree of digital self-sufficiency and control over data within their jurisdiction, typically justified with reference to national security.

• State-based model

The state-based model is characterised by a blurred distinction between public and private sectors, where digital sovereignty is conceived as a tool for nation-building and economic growth. States seek digital self-sufficiency by restricting foreign companies and supporting domestic digital champions, especially in emerging technologies. The state plays a key role through investment and strategy and is more active (compared to the centralisation model) in exporting its technical standards and governance norms. Once established, national digital operators are sent into the international arena to compete and spread influence.

Fratini et al categorise the approach taken by nation states according to these models as follows:

Figure 2. Different approaches to digital sovereignty taken by nation states

As illustrated by figure 2, countries such as the US have arguably pursued a more market-orientated and competition approach to digital sovereignty whilst the European Union has opted for a model based more on rights and state controls. The UK’s position is explored in depth below, but Fratini et al contend that it and nations like Canada, New Zealand and Australia have pursued an approach that is more balanced between state control and competition. All of these nations have taken a softer approach to regulation compared with nations such as China and Russia.

4. How has the UK responded?

The UK has not pursued an overarching strategy of digital and technological sovereignty. Rather, as set out in the government’s ‘Modern industrial strategy’ from June 2025, and the accompanying ‘Digital and technologies sector plan’, ministers intend to build “sovereign capability” in key emerging digital technologies: advanced connectivity; AI; cyber security; engineering biology; quantum technologies; and semiconductors.^[37]

This ‘sovereign capability’ does not mean digital sovereignty in the sense of self-sufficiency in these areas. The government’s approach is to build on the UK’s existing strengths (especially where there are national security applications), address risks to supply chains through international partnerships, and shape technological development by influencing international standards in multilateral forums.^[38]

Some of the key developments and challenges in UK digital and technology policy in the context of this debate are examined below.

4.1 Digital services tax

There have been efforts to reform global systems of taxation to take account of the increased digitisation of products and services for many years. In response to some of the challenges posed by the digitalisation of the economy to the international corporation tax system, in April 2020 the UK implemented the digital services tax (DST). DST is a 2 percent tax on certain revenues to ensure that providers of search engines, social media platforms, and online marketplaces pay UK tax on their digital services to reflect that they derive value from user-related activities in the UK. The tax applies to groups with over £500mn global and £25mn UK revenues from the in-scope activities.^[39]

As noted by the Treasury, DST was “always intended to be an interim tax measure”, which will be replaced once an “appropriate global solution” is in place.^[40]

Efforts to find such a global solution have predominately been focused on the Organisation for Economic Co-operation and Development (OECD)’s ‘base erosion and profit shifting’ or ‘BEPS’ initiative, and on plans published in 2019 based on the following two ‘pillars’:^[41]

• pillar 1: proposals for determining where tax should be paid and on what basis (‘nexus’), as well as what portion of profits could or should be taxed in the jurisdictions where clients or users are located (‘profit allocation’)
• pillar 2: a proposal for a global minimum corporate tax level

In October 2021, more than 135 countries, including the UK, reached an agreement on pillar 2. The agreement, also known as the OECD/G20 Inclusive Framework on BEPS (Inclusive Framework), stipulates that the largest corporate groups will be subject to a minimum rate of tax, by enacting a global minimum corporate tax rate of 15%.^[42] As part of the agreement, the government has introduced two new taxes in the UK: a domestic top-up tax, and a multinational top-up tax. These taxes apply to accounting periods beginning on or after 31 December 2023.

However, one of President Trump’s first acts in his second term was to issue an executive order rejecting the application of the pillar 2 deal in the United States.^[43] Instead, the 147 jurisdictions which have now agreed the OECD/G20 Inclusive Framework have agreed on a tax package establishing a so-called “side-by-side” mechanism.^[44] US Treasury Secretary Scott Bessent has called the updated agreement a “historic victory preserving US tax sovereignty”.^[45] Yet Pascal Saint-Amans, project lead on the EU Tax compass at the Bruegel thinktank, suggests the reality is more nuanced:

[…] the agreement does not amount to a full dismantling of pillar 2. It preserves the minimum tax architecture. Differentiated treatment for the US is regrettable but the US has its own minimum tax, which, with a 14 percent rate, is only slightly below the global rate but which applies a much narrower base. Reality is therefore more nuanced than US victory claims, which may explain why all IF [Inclusive Framework] members, even China, India and developing countries more broadly have signed the new agreement, despite reservations about an asymmetrical deal twisted in favour of the US.^[46]

Reaching consensus on pillar 1 has proved even more difficult. In particular, as noted above, the US government has argued that individual digital service taxes—such as that introduced by the UK and which would fall under pillar 1—discriminate against American technology companies.^[47] The US government has threatened retaliatory action in the form of tariffs.

The UK Treasury contends that “significant progress has been made over the past years in negotiations on the pillar 1 solution, even if a final agreement has not been reached”.^[48]

The Treasury said the implementation of the DST had cost HMRC £6.3mn. The DST raised around £800mn in revenues in 2024–25. DST revenues by year to date are set out below:

(HM Treasury, ‘Digital services tax review: Report’, November 2025, p 12.)

4.2 A failure to scale?

One of the difficulties in developing digital sovereignty is an absence of domestic providers of key services and innovations. The House of Lords Science and Technology Committee is among those who have pointed out promising science and technology companies have chosen, for a variety of reasons, to scale overseas rather than in the UK. Reporting in November 2025, the committee argued this issue had reached a “crisis point” which was causing the UK economy to “bleed out”.^[49] The committee said:

The “failure to scale” for UK technology companies is not new. It has been a deep-rooted problem. But we are in a new technological and geopolitical context, with the rise of AI and the fracturing of the old global order. Technological sovereignty—who owns and operates critical technologies—is more important than ever. Domestically, the UK has seen sluggish productivity growth and near-flat real wages since the global financial crisis. The UK economy is simply not working, and the consequences are easy to see.

This has now reached a crisis point. The UK has seen a procession of promising science and technology companies moving overseas rather than scaling here. The UK’s inability to retain the economic benefits of its R&D endeavour is a fatal flaw in any growth strategy.^[50]

The committee highlighted the UK’s competitive research base but said this was failing to translate into successful companies:

The UK has an internationally competitive research base, especially in its universities, albeit this is now under financial threat. There is a healthy and growing scene of start-ups and spin-outs; great progress has been made at their early stages. But the UK has been less effective both in scaling up companies that start in the UK and in diffusing new science and technologies across the economy. The result is a country with four of the top ten global universities, but only three of the top 100 industrial R&D spenders globally and none in the top ten.^[51]

The committee argued that the UK risked becoming an “incubator economy” whose young, fast-growing companies continually move abroad, “taking much of the economic benefit in jobs and tax receipts with them”.^[52] The committee made several recommendations including clearer leadership from government, efforts to unlock institutional investors, and reforms to public procurement, including a mandatory target for departments to spend with innovative UK based small and medium enterprises.

In its response, the government argued that the UK was one of the “best places to start a science and technology company in the world”.^[53] It pointed out:

We are home to four of the world’s top ten universities and our venture capital market is the third largest globally, behind only the US and China and leading in Europe, with over £17bn invested in innovative UK companies last year. We are world-leading in many technology sectors such as quantum, where we have the second highest number of quantum start-ups globally, and the sector has attracted the second highest volume of private investment globally.^[54]

However, the government said it recognised “the urgency” of going further and improving support for science and technology companies to scale in the UK, “something that has been a longstanding problem for the UK over several decades”.^[55] It said that it was committed to an integrated approach that prioritised the right support at the right times:

To deliver the vision set out in the industrial strategy, the government is taking an integrated approach to supporting innovative companies, making sure companies can access the right support at different points in their growth journey, with funding complemented by access to other initiatives from across government on procurement, skills, infrastructure, and regulation. The government’s science and technology framework sets out how the government intends to take an integrated, systems-level approach to science and technology policy in order to deliver on its national missions.^[56]

It added:

The government is also driving forward a cross-department approach to supporting scaleups, including tech companies, which will be rolled out by mid-2026. This will focus on identifying and unblocking barriers to growth at pace, including procurement, access to finance and regulation, and providing easier access for companies into government. As part of this, DSIT [the Department for Science, Innovation and Technology] will take a proactive approach to identifying strategically important frontier tech companies across its growth, investment and sovereignty agendas and will actively engage with these companies to support them to stay, scale and invest in the UK.^[57]

The government contested that this was underpinned by record investment in research and development:

Through the £86bn R&D investment in the spending review and industrial strategy, we are backing founders and innovators and prioritising the sectors with the highest potential for growth. Furthermore, business expenditure on R&D is strong, increasing by 5% to £55.6bn, the highest on record under the updated ONS methodology. Central to delivering this will be reform of UK Research and Investment, which includes Innovate UK. Under new leadership, they will ensure that they support not just world-leading R&D but also identify and grow high potential businesses which have the potential to be world leaders, investing £7.4bn over the next 4 years to help companies scale and grow.^[58]

4.3 Sovereign AI capabilities

AI is an area where the government has been more explicit in talking about sovereignty. In the government’s AI opportunities action plan, published in January 2025, ministers said they intended the UK to be an AI ‘maker’ not a ‘taker’:

• Position the UK to be an AI maker, not an AI taker:As the technology becomes more powerful, we should be the best state partner to those building frontier AI. The UK should aim to have true national champions at critical layers of the AI stack so that the UK benefits economically from AI advancement and has influence on future AI’s values, safety and governance.^[59]

The minister for AI, Kanishka Narayan, expanded on this view in evidence to the Joint Committee on Human Rights in February 2026, defining AI sovereignty as:

[…] the ability for a state to have strategic leverage when it comes to this technology, such that it can ensure ongoing access to critical inputs, and ongoing assurance that its wider economic and national security objectives can be met more broadly.^[60]

He said that the UK could gain strategic leverage through three important steps. The first was to have “enough of the critical inputs” required for AI development and usage, particularly sophisticated microchips. The second was “to make sure that you have some diversification in who you are procuring critical inputs from so that you can bargain effectively”. To that end, he said the government had been working “with a series of companies in every part of the stack so that we were able to build some diversity into the landscape”. He said the third ‘rung on the ladder’ was “ultimately to build British”.

As announced in the AI opportunities action plan, the UK has set up the Sovereign AI Unit which has been “specifically designed to build homegrown capability in emerging areas of the AI ecosystem”.^[61] Backed by £500mn in funding, the government intends that the unit will strengthen the UK’s AI capabilities by:

• Investing in UK companies to support AI national champions. The Sovereign AI Unit will work with Innovate UK, and the British Business Bank to support high-potential start-ups to start and scale their companies in the UK.
• Creating and developing UK AI assets and enablers such as data, compute and talent.
• Making the UK the partner of choice for frontier AI companies. This will ensure that the UK public and private sectors benefit from reliable access to, and influence over the development of, cutting edge technology, ensuring everyone across the country benefits from transformative AI.^[62]

The government contends the unit has started delivering, with projects including:

• OpenBind: £8mn seed investment in a Diamond Light Source (DLS)-led consortium to create the world’s largest open protein-ligand dataset and secure sovereign leadership in AI-native drug discovery.
• Partnerships: with Anthropic and Cohere to gain a UK “stake” in frontier AI development, ensuring that strategically important companies are invested in the UK and that their presence is secure.
• Encode Fellowship: seeking to expand a flagship one-year, entrepreneurially focused program run by Pillar VC and ARIA aiming to get top global AI talent working in UK labs on various domains in AI for Science within four months.^[63]

In June 2026, the government also released its AI hardware plan.^[64] In it, ministers noted that while global compute—the computing processing power used to train and deploy AI models—remains concentrated, a more diversified ecosystem is emerging:

The AI stack is evolving quickly. While global compute remains highly concentrated, the hardware that underpins it is diversifying. Different AI workloads—from training and inference to edge and embodied AI—increasingly require specialised architectures and system designs. This reflects a shift from centralised model training towards more distributed and application-specific computing. This shift is creating new opportunities across the stack, enabling a more diverse and resilient ecosystem and opening space for new entrants. It also reinforces the need for AI systems that are secure by design, with hardware playing a critical role in enabling safe and trustworthy adoption at scale.^[65]

The government contends that this has created opportunities for the UK:

This is a major economic opportunity. AI hardware—the semiconductor technologies that underpin AI—is one of the fastest-growing segments of the global economy, central to productivity growth, high-value job creation and economic resilience. Capturing value in this space will be critical to the UK’s long-term position in AI and its role in global supply chains.

The UK is well placed to lead in the future hardware technologies required for AI. It combines world-leading strengths in chip design and IP, a growing base of innovative AI hardware firms, and differentiated capability in enabling technologies such as photonics and advanced materials. These strengths are increasingly co-locating in practice—with the UK emerging as a major hub for frontier AI labs—creating a unique opportunity to bring hardware, models and real-world workloads together.^[66]

As part of the plan, the government has announced £1.1bn of public and private investment which aims to convert those strengths into scaled capability and build globally competitive AI hardware companies in the UK. This includes attracting and anchoring international investment; maximising the adoption of UK-developed technologies across AI systems; and establishing the UK as one of the “leading places in the world for next-generation AI hardware”.

The government’s commitments to distribute that £1.1bn across the areas of innovation, skills, procurement, and investment can be summarised as follows:^[67]

Innovation 

• A £120mn investment in AI hardware innovation including:
  • an AI hardware innovation programme, designed in partnership with UKRI (UK Research and Innovation) and the UK Semiconductor Centre.
  • an expanded Scaling Inference Lab, with additional funding of at least £20mn to support more companies and larger-scale system builds for evaluation, delivered in partnership with ARIA and CommonAI and building on ARIA’s existing £50mn investment in this area.
• A new £18mn hardware security R&D programme, strengthening the secure deployment and scaling of AI hardware.
• A refocused Semiconductor Catapult, building on the existing Compound Semiconductor Applications Catapult to focus on AI hardware and support UK-based companies.
• A trusted network of manufacturing and foundry partners, built through the UK Semiconductor Centre, providing UK innovators with access to a clear pathway from research, design and prototyping to production and scale.

Skills 

• An £80mn investment in semiconductor and AI hardware skills, including:
  • an expansion of the semiconductor skills programme to £48mn, delivering increased undergraduate bursaries in subjects such as electrical and electronic engineering and materials science.
  • a new £12mn Centre for Doctoral Training (CDT) in Chip Design, strengthening the long-term talent pipeline in chip design and AI hardware.
  • a new £20mn targeted investment of the TechFirst programme, supporting 500 more UK PhD students to access top up funding and support throughout their PhD.
• Work with employers to identify pathways into semiconductor and AI hardware careers, led by Skills England, DWP [the Department for Work and Pensions] and the UK Semiconductor Centre
• A strategic industry partnership with Armthrough TechFirst, supporting the development of the UK’s future AI hardware workforce.

Procurement 

• A new £750mn heterogeneous AI supercomputer for the AI Research Resource (AIRR), enabling different types of advanced compute technologies, including novel AI architectures and, over time, quantum computing, to be integrated within a single system and used on real research workloads.
• A £400mn procurement opportunity for specialised chipswithin this £750mn AIRR heterogeneous supercomputer programme.
• A new digital procurement model through the multi-billion pound National Cloud Infrastructure Programme (NCIP), providing opportunities for UK AI hardware companies to deploy at scale in cloud infrastructure and demonstrate readiness for global markets.

Investment 

• A new deeptech hardware venture fund, led by Playground Global, backed by up to £150mn from the British Business Bank.
• Support for AI hardware companies through the £500mn Sovereign AI Fund, which is prioritising compute.
• Support for UK AI hardware companies to access financing through UK Export Finance (UKEF), enabling firms to scale internationally.
• Support for UK AI hardware companies to scale internationally by navigating export requirements.

The government has also published ‘Shaping tomorrow: The UK’s digital standards strategy (2026 to 2030)’, which sets out how the UK intends to leverage the economic potential of digital standards and support the growth of its tech industry to “reinforce our leadership in a digitally connected world”.^[68]

Further, on the question of foreign ownership, in answer to a recent parliamentary question on what assessment the government has made of the level of dependency on overseas AI providers, Kanishka Narayan said:

The UK recognises the risks that come with highly concentrated markets and the need to strengthen resilience in critical technologies. We see this less as an issue of dependence on any single country, but rather an overreliance on narrow supply chains or single points of failure. Our approach is therefore about diversification, trusted interdependence and capability-building with like-minded partners and allies.

We will continue to use the best technology and welcome inward investment because that is what our public services and economy demand. We have made a decisive move towards backing more British AI companies, especially in areas where the UK can build and defend comparative advantage and where government can add value. We are investing £1.6bn in the UK’s sovereign AI capabilities, with £500mn invested in UK AI firms via our sovereign AI fund, and £1.1bn via our AI hardware plan.^[69]

The use of copyrighted material to train certain large language AI models remains a contested issue. Baroness Kidron has questioned ministers on whether the UK’s sovereign AI fund will invest in or support companies that train models on copyrighted work without a licence. In response, Department for Science, Innovation and Technology Minister Baroness Lloyd of Effra said:

[…] We have been clear that copyright rules should be respected and the use of copyright works to train AI in the UK requires a licence unless an exception applies. Companies supported by the sovereign AI fund are expected to comply with applicable UK law, including copyright. When we are talking about compliance in relation to grant-funded compute allocations, they equally must comply with copyright law while undertaking that funded activity.^[70]

However, writing in the Times in April 2026, Baroness Kidron called assurances that investments will comply with UK law as “disingenuous”, suggesting that there were already indications that some firms receiving investment, or valuable access to public compute, are based in Silicon Valley and are refusing to disclose if the data sets used to train their models are consistent with UK copyright law.^[71]

4.4 Stakeholder views and calls for a digital sovereignty strategy

Several parliamentarians and stakeholder organisations have called upon the government to create a more explicit and cross-sector digital sovereignty strategy for the UK. It was the subject of an early day motion in the House of Commons in January 2026,^[72] and has also been discussed in the context of the Cyber Security and Resilience (Network and Information Systems) Bill as it progresses through Parliament.^[73]

A report by the House of Commons Business and Trade Committee on economic security published in November 2025,^[74] also called upon the government to take a “more structured” approach to identifying areas where sovereign capability is needed. In its response, the government said:

The government also recognises the need to mitigate the risk of being overly dependent on foreign suppliers and so the national security strategy emphasises the need to nurture and protect these high-growth sectors from harm, to develop asymmetric strengths, and to enhance the UK’s geopolitical leverage.^[75]

The committee was critical of the government’s response, suggesting other nations had gone much further in articulating risks, dependencies, and thus targeted policy interventions:

Our allies in the EU, Japan and the United States are all creating a more disciplined model, characterised by much clearer articulation of risks, strategic dependencies and targeted policy interventions. The government response does not set out an equivalent prioritisation for the UK, nor does it identify which dependencies are unacceptable and must be reduced.^[76]

The House of Commons Science, Innovation and Technology Committee has also called on the government to do more on technology sovereignty.^[77] It concluded:

Setting a cross-public sector definition of sovereignty and agreeing a strategy that is clear about the sectors and services where sovereign capability matters most, is a prerequisite for the effective use of technology by government departments and public bodies. We are therefore concerned by the current lack of clarity over how the government is approaching this important topic. Failure to agree and enact a clear sovereignty strategy is not only a barrier to the successful deployment of technology, but leaves the UK at the mercy of foreign commercial and state actors that do not share our strategic interests.^[78]

In its written evidence to the Cyber Security and Resilience Public Bill Committee, the Open Rights group, which advocates for digital rights and freedoms, suggested the bill failed to address the cyber security risks posed by the actions of overseas vendors of software or hardware in response to requests from their national governments.^[79] Yet the group also said such risks went beyond cyber security and needed to be considered in a holistic manner:

Digital sovereignty risks are broader than cybersecurity, including risks to economic development; and risks to foreign policy flexibility, from the mere threat of cybersecurity risks. It is a strategic question that needs to be considered in a broad manner.

[…]

In the tech sectoral plan, this bill is mentioned as part of its industrial strategy, and develops the government’s approach to AI, but focuses on private sector growth without fully developing the theme of “upheaval in geopolitics” and how the UK might “adapt, compete and endure”; the missing step is to ask how the UK can secure a sovereign infrastructure, and whether government spending is in fact working against the IT sector and UK economy’s interests, as well avoiding discussion of the cyber security risks that result.^[80]

The Open Rights Group added:

Managing digital sovereignty is an active process in the EU and key European countries, but remains neglected in the UK. At this stage, Parliament does not need to dictate the details of a digital sovereignty strategy, which would need to cover many more issues than there is space to detail here. However, the bill offers Parliament the opportunity to recognise that cybersecurity and sovereignty risks from vendor dependency exist, and to establish a duty for government to put a digital sovereignty strategy in place to kick start a process of change.^[81]

In an April 2026 report for the BT Group, the consultancy firm Assembly also suggested greater digital sovereignty could unlock significant opportunities and outlined a range of actions the government could take further to this aim:

Our estimates show that the adoption of digital sovereignty could unlock significant economic benefits for the UK by enabling businesses to realise AI’s full potential, encouraging investment into local data centres, protecting against regulatory fines, growing a domestic sovereign cloud market, and limiting the economic costs of data, IP and asset loss.

[…] There are four actions the UK government could take to support the adoption of secure and reliable UK-based digital services: i) outline a framework for sovereignty; ii) highlight sectoral requirements for sovereignty; iii) use natural convening powers to connect the supply and demand sides of the market; and iv) revise procurement rules and practices.^[82]

The Tony Blair Institute for Global Change has suggested that no country can be entirely self-sufficient in the AI era:

Developing and deploying frontier AI requires enormous resources: billions of dollars in compute, data and engineering talent, alongside hyperscale data centres and cutting-edge semiconductors. These capabilities are overwhelmingly concentrated in the United States and China, which together control more than 90 per cent of global AI data-centre capacity. Most states will never be able to build or sustain frontier AI infrastructure on their own.

The dependencies this creates are increasingly seen as either strategic vulnerabilities or levers of geopolitical influence. Countries are left to confront a choice about how to remain competitive in the AI era: join the resource-intensive race to train the world’s most advanced models or focus on deploying AI at scale, bolstering competitiveness through deployment rather than domestic frontier systems.^[83]

In this context, it contends that thinking of sovereignty in terms of bolstering domestic capacity and enforcing stricter regulations could be counterproductive:

[…] calls for “AI sovereignty” are emerging, presented as an imperative to exercise exclusive control over frontier technology and its use, to build domestic capabilities and to enforce stricter regulations on foreign technology. Although motivated by legitimate economic and security concerns, this instinct reflects a narrow and ultimately counterproductive understanding of sovereignty: one that equates autonomy with full technological control and treats interdependence as a vulnerability to be eliminated.

A more realistic and grounded understanding of sovereignty that reflects the realities of interdependence rather than the illusion of isolation is therefore needed. Sovereignty in the age of AI is not a binary condition to be achieved or lost. It is fundamentally a question of agency and choice—the ability of a state to make deliberate, future-oriented decisions about how AI is integrated, governed and used in line with its national goals.^[84]

The institute argues that the response should be grounded in domestic capabilities but cannot be achieved in isolation:

Effective AI sovereignty therefore cannot be pursued through isolation. It must be deliberately negotiated. Governments will need to cultivate domestic strengths where they matter most, secure predictable access to frontier capabilities, design partnerships that preserve flexibility, and invest in the institutions and talent required to evaluate, govern and adapt AI systems. Above all, national competitiveness in the AI era will depend as much on deploying AI widely across the economy and public sector as on building frontier models themselves. For many countries, this path will deliver far greater returns than entering the resource-intensive race to train the world’s most advanced systems.^[85]

A subsequent report from the institute report argued that ‘middle powers’ such as the UK cannot expect to compete with the US and China on frontier AI model development, and should focus on supporting open-source models, services, and data.^[86]

Others have similarly argued for more emphasis on open-source technologies. For example, the Open Rights Group contends:

The ability to control, maintain and secure government systems is immediately compromised when government does not have access to source code, cannot modify software for itself, and does not have the long-term rights to modify the software. Consequently, the most favourable systems are open source, which allow government to redeploy, modify the code, replace one vendor for another, etc. Open source ensures full autonomy and complete ability to risk manage, where proprietary systems require a loss of autonomy and dependence on specific vendors.^[87]

Similarly, OpenUK has also called on the government to create a national foundation to support open source development and open source-based businesses.^[88]

In a blog post for TechUK, Nick Roberts, director of sovereign cloud at Rackspace Technology, said that much of the debate over digital sovereignty had centred on where data lived. However, he said whilst that question was still important, decision rights must also be considered:

Who controls the infrastructure? Who operates it day to day? Who can access, audit and intervene when systems are under pressure? And under which governance frameworks do those decisions sit? If digital sovereignty is to deliver real value for the UK, it must answer those questions clearly and consistently.^[89]

Mr Roberts said that required more than just UK-based data centres; it “demands jurisdictional clarity across operations, UK-cleared personnel where appropriate, and governance models that ensure control always remains within defined jurisdictional boundaries”. He also argued that digital sovereignty was intertwined with national resilience:

Recent global disruptions have exposed how tightly coupled digital systems are to international events. Infrastructure supporting essential services must be designed not just for efficiency, but for continuity under pressure.

A digitally sovereign approach should enable:

• resilient architectures across UK regions
• clear escalation and crisis-management paths within national jurisdiction
• operational independence when global systems are disrupted^[90]

He said questions over digital sovereignty took on “new urgency” in the age of AI. He argued:

Digital sovereignty is no longer a procurement checkbox or a compliance footnote. It is an operating model. For the UK, the next phase is about moving beyond data residency toward decision rights: control, transparency and resilience designed into the fabric of digital infrastructure. That shift is essential if the UK is to adopt AI responsibly, protect critical services and remain confident in how its digital systems are governed in an increasingly complex world.^[91]

5. The European approach

There is more emphasis on an over-arching strategy for digital sovereignty in the European Union in comparison with the UK. After several years of increasing emphasis on developing sovereign digital and tech capacity, on 3 June 2026 the European Commission adopted a set of measures to bolster the EU’s digital autonomy in what it described as a “major shift in its approach to technology”.^[92] The proposal followed the 2025 ‘State of the Union’ address by Commission President Ursula von der Leyen, where she described Europe’s technological sovereignty as key to strengthening competitiveness, resilience, and strategic autonomy in a fast-changing digital world.^[93]

The package includes two legislative proposals:

• The Chips Act 2.0

The Commission has adopted a proposal for the Chips Act 2.0, which introduces new measures to further boost the chips industry, reduce strategic dependencies and support advanced chip production in the EU.

The EU remains dependent on third countries in key areas such as advanced chip manufacturing or semiconductor design. Securing a stable chips supply is necessary to ensure that critical infrastructures and technologies remain secure, resilient and aligned with European values.

The Chips Act 2.0 builds on the progress made by the original Chips Act and will both reinforce current European strengths (including mainstream chips) and build capacity in cutting-edge semiconductor technologies. This will allow the EU to maintain its position as an indispensable player in the value chain, while strengthening its resilience, and reducing strategic dependencies and supply chain vulnerabilities.^[94]

• Cloud and AI Development Act (CADA)^[95]

The commission contends that the CADA will reinforce energy-efficient data centre capacity, while complementing the ‘Apply AI strategy’ to boost the adoption of artificial intelligence and cloud across Europe.

It will focus on three objectives:

1. Research, development and innovation: supporting efforts to deploy the next generation of cutting-edge and sustainable cloud and AI technologies
2. Capacity: accelerating the conditions for deploying data centres across the EU, focusing on facilities that enhance essential public sector functions
3. Autonomy: introducing a single EU-wide assessment framework for cloud and AI sovereignty, accompanied by a public sector adoption mechanism

The package also includes an ‘EU open source strategy’ and a ‘Strategic roadmap for digitalisation and AI in energy’. According to the Commission, the EU open source strategy aims to place open source at the centre of the EU’s technological sovereignty by “promoting European open alternatives to non-EU proprietary solutions in critical domains”.^[96] The EU open source strategy aims to strengthen Europe’s open digital ecosystems by supporting the development, scaling, deployment and long-term sustainability of open source technologies across both the public and private sectors. Meanwhile, the strategic roadmap for digitalisation and AI in energy aims to accelerate the rollout of digital solutions, including European sovereign AI solutions, in areas that are important for the decarbonisation process, such as electricity grid optimisation, energy efficiency in buildings and industry, and demand-side flexibility.^[97]

Noting that the European Union currently relies on non-EU countries for over 80% of key digital products, services, infrastructure, and intellectual property, the Commission contends that this new package will enable Europe to:

• strengthen its competitiveness, resilience and security
• ensure strategic autonomy in key digital technologies
• support open and fair digital markets
• protect citizens’ rights and democratic processes in the online space
• enable innovation and long-term technological leadership^[98]

Speaking to the proposals, Commission President Ursula von der Leyen said:

We cannot afford to depend on others for the technologies that keep our hospitals running, our energy grids stable and our services secure. This is about protecting our citizens, defending our interests and making our own choices. Europe has the talent, the research excellence, the industrial base and the single market. Together, we must turn these strengths into technological sovereignty.^[99]

---

Image by NASA on Unsplash